ComfyUI Node Reference

Combines AnimateDiff frames into video or GIF output.

Injects a motion model for coherent animation generation.

Applies ControlNet guidance from a reference image.

Encodes a text prompt into conditioning for the sampler.

Loads dual CLIP encoders for SDXL and Flux models.

Creates a blank latent canvas for txt2img generation.

Runs the diffusion process to generate images from noise.

Extended sampler with start/end step and noise control.

Loads a checkpoint and outputs MODEL, CLIP, and VAE.

Loads a ControlNet model for image-guided generation.

Loads an image from disk for img2img or ControlNet.

Applies a LoRA weight file to a model and CLIP encoder.

Loads a standalone VAE to override the checkpoint default.

Loads a video and extracts frames for processing.

Saves images to PNG with embedded workflow metadata.

Resizes an image using configurable interpolation.

Resizes latent for hi-res fix and multi-pass workflows.

Converts latent output into a visible pixel image.

Encodes a pixel image into latent space for img2img.

Combines image frames into MP4, WebM, or GIF video.